#!/usr/bin/env ruby
#
# Slugalizer 0.1.9
#
# http://termos.vemod.net/slugalizer
#
# Released into the public domain
#
# Please send bug reports and improvements
# to christoffer.sawicki@gmail.com
#

require "active_support/multibyte"

module Slugalizer
  extend self
  
  class Error < ::StandardError
  end
  
  WORD_SEPARATORS = %w<- _ +>
  
  def slugalize(text, word_separator = "-")
    unless WORD_SEPARATORS.include?(word_separator)
      raise Error, "Word separator must be one of #{WORD_SEPARATORS.join(' ')}"
    end

    ActiveSupport::Multibyte::Handlers::UTF8Handler.normalize(text.to_s, :kd).
      gsub(/[^\w\s\-\+\.]/n, "").
      strip.
      gsub(/\s+/, word_separator).
      downcase
  end  
end

if __FILE__ == $0
  require "test/unit"
  
  class SlugalizerTest < Test::Unit::TestCase
    def assert_slug(expected_slug, *args)
      assert_equal(expected_slug, Slugalizer.slugalize(*args))
    end
    
    def test_converting_to_string
      assert_slug("", nil)
      assert_slug("1", 1)
    end
    
    def test_identity
      assert_slug("abc.1_2_3-4", "abc.1_2_3-4")
    end
    
    def test_asciification
      assert_slug("raksmorgas", "räksmörgås")
    end
    
    def test_downcasing
      assert_slug("raksmorgas", "RÄKSMÖRGÅS")
    end
    
    def test_special_characters
      assert_slug("raksmorgas", "räksmörgås!?")
    end
    
    def test_accented_characters
      assert_slug("acegiklnuo", "āčēģīķļņūö")
    end
    
    def test_chinese_text
      assert_slug("chinese-text", "chinese 中文測試 text")
    end
    
    def test_stripped_character_then_whitespace
      assert_slug("abc", "! abc !")
    end
      
    def test_single_whitescape
      assert_slug("smorgasbord-e-gott", "smörgåsbord é gott")
    end
    
    def test_surrounding_whitescape
      assert_slug("smorgasbord-e-gott", " smörgåsbord é gott ")
    end
    
    def test_excessive_whitescape
      assert_slug("smorgasbord-ar-gott", "smörgåsbord  \n  är  \t   gott")
    end
    
    def test_word_separator_parameter
      assert_slug("smorgasbord-ar-gott", "smörgåsbord är gott", "-")
      assert_slug("smorgasbord_ar_gott", "smörgåsbord är gott", "_")
      assert_slug("smorgasbord+ar+gott", "smörgåsbord är gott", "+")
    end
    
    def test_invalid_word_separator
      assert_raise(Slugalizer::Error) do
        Slugalizer.slugalize("smörgåsbord är gott", "@")
      end
    end
    
    def test_handling_of_word_separator_chars
      assert_slug("abc_-_1_2_3", "abc - 1_2_3", "_")
    end
    
    def test_handling_of_stuff
      assert_slug("foo-+-b_a_r", "foo + b_a_r")
    end
    
    def test_with_kcode_set_to_utf_8
      old_kcode = $KCODE
      $KCODE = "u"
      assert_slug("raksmorgas", "räksmörgås")
    ensure
      $KCODE = old_kcode
    end
  end
end